The operation of radars and laser radars generally is thought of in terms of (a) sending out a pulse and timing the return signal to obtain target range or (b) sending out a wave of constant frequency and measuring the frequency of the return signal to obtain the doppler shift and thereby target velocity. Other waveform modulations also have useful applications. One such waveform modulation is linear frequency modulation (FM), where the frequency of the radar is linearly increased or decreased with time; this is known generally as "FM-chirp." The frequency difference between the target return at any instant and the frequency of the transmitter at the same instant (measured by heterodyning the return with a sample of the transmitter) is related to both the range and velocity of the target. Target velocity affects the frequency difference through the doppler shift; target range affects the frequency difference through the change in transmitter frequency during the round trip transit time. Available methods for separating the effects of range and velocity generally rely on reversing or eliminating the chirp slope for some period of time. There are also applications where there is no need to separate these effects. One such application is when the target travels so slowly that the doppler is small or negligible. Another application is for imaging where the entire target has the same velocity, so that changes in frequency represent relative changes in range. For these applications, a repetitive, one directional chirp is the waveform of choice, and system performance depends on the linearity and repeatability of the chirp and the amount of power transmitted. This invention is a low-loss method of generating linear, repeatable, one-directional chirps from a laser transmitter of an optical radar system.
In the prior art, chirping is achieved by employing an electro-optic crystal modulator within the laser cavity (resonator). The laser optical frequency is modulated when the effective optical length of the crystal is varied in response to a suitably shaped electrical pulse applied to electrodes situated at either side of the crystal. A typical electro-optic crystal is formed from cadmium telluride, 5 cm long, 2 mm by 2 mm in cross section, and highly polished on all surfaces.
Many difficulties are encountered with this prior art approach: The crystals are both expensive and fragile. Optical alignment of such crystals is difficult because of their long, narrow shape. Some inherent characteristics of the crystal itself also affect performance. The imperfect optical quality of such crystals leads to optical distortions and birefringence. Strain-optic effects cause transverse variations in the index of refraction that distort the optical beam and degrade beam quality. Piezo-electric effects produce dimensional changes that generate acoustic waves that might fracture the crystal or introduce additional strain-optic effects. Thermal-optic effects result from absorbed optical energy. Because this absorption is not distributed uniformly throughout the crystal, it causes mechanical strains and index of refraction gradients. Typical insertion loss resulting from placing such a crystal into the cavity is about a ten to twelve percent round trip loss.
To operate effectively, a high voltage ramp (0 to 1600 volts) swept at periods of 3 to 50 microseconds must be applied. Arcing and flashover either across the crystal or to adjacent elements must be avoided. As will be explained subsequently, excessive optical length introduced into the laser cavity by such a crystal significantly limits the maximum total range of chirp modulation that can be achieved.
Chirp linearity and repeatability limit the resolution and signal-to-noise ratio of the radar system. Linearity of a system employing an electrooptic crystal depends mainly on linearity of the voltage ramp as a function of time plus dynamic disturbances that alter the index of refraction. These disturbances result from piezo-electric and strainoptic effects. Birefringence and gradients in the index of refraction induced by heat and strain also degrade linearity of the optical frequency chirp. Similarly, repeatability of the chirp slope depends on the repeatability of the voltage drive, dynamic strain-optic effects, and temperature dependence of the electro-optic coefficient. With crystal modulators, deviations from linearity of the chirp ramp can be held to one percent. Chirp slope can be held constant to within one percent from chirp to chirp.
The use of this invention can improve chirp slope linearity by about two orders of magnitude. Chirp slope repeatability will be about the same as in the prior art. It is important to note, however, that variations in slope repeatability resulting from the use of this invention are systematic and known, while variations of the prior art are essentially unpredictable. In many cases, known and systematic variations can be compensated for, which will yield laser radar performance improvements similar to what could be obtained from improving slope repeatability by one or two orders of magnitude.